Overview
RenalDB is a tool designed to assist researchers for the in silico screening of
enriched/specific transcripts of humans, mice, and zebrafish with respect to nephrotic tissues and cells,
developmental stages, and other metadata. Furthermore RenalDB serves as a test case for future databases
focused on multi-attribute representations of RNA expression. The objective of RenalDB is
simple; that is, to make a hypothesis-driven research to be easier. We built RenalDB from
the standpoint of researchers to answer very fundamental questions in nephrology, such as, “Are
there any RNAs (transcripts) that are expressed specifically in patients suffering from kidney
diseases but not in healthy people?”
Search
While all types of RNA transcripts are included in RenalDB, the primary motivation for its creation is the study of lncRNAs.
One of the major features of lncRNAs is high specificity of their expression patterns to a certain cell and/or tissue.
To highlight this feature, RenalDB features advanced search capabilities.
The search bar allows users to combine a variety of tags with boolean operators to create
arbitrarily complex searches. See the table under "Queries" for a list of supported search
tags and boolean operators.
Search View
Clicking on the [LOCI] button displayed at the top of the window will take the users to the loci view. This view contains a search form followed by a list of relevant "loci" (i.e. genes, transcripts, etc.). If no query is entered, the list will contain the set of all loci in RenalDB. When the [GO] button is pressed, a search will be executed, and the filtered list of loci will be returned. Detailed information about building queries can be found by clicking the button with a question mark, or on this page. Refer to the table under "Queries" below. The number of rows can be changes the value shown in the "Rows Per Page" field using the drop-down list. The page of the results can be changes by modifying the number next to "Page" or by clicking the "Previous" or "Next" buttons.Universal Genomic Accessions
The users may notice the unconventional looking accession numbers used in RenalDB. These are Universal Genomic Accessions (UGAs) a hash-based accession system that allows for the de-centralized accession system. For details about UGAHash, please refer to the UGAHash web interace (http://ugahash.uni-frankfurt.de) and the following publication:For more information regarding UGAs see the UGAHash website.
Or:
Weirick, T., John, D., Uchida, S. (2016). Resolving the problem of multiple accessions of the same transcript deposited across various public databases. Briefings in bioinformatics, bbv067. PMID: 26921280
UGAs have a number of advantages over traditional accession systems. In RenalDB, the main advantage is easy retrieval of relations to other genomic databases (e.g. ENSEMBL, UCSC Genome Browser). These relations are handled via CORS (cross-origin resource sharing) by requesting to the UGAHash server. The relations are updated automatically whenever the UGAHash server is updated. From the search bar, the users can use accessions from other databases (e.g. ENSEMBL, NCBI, NONCODE).
Searchable Values
A number of the search tags will only return useful results if given specific tags. The tables below contain the names that can be used for queries.Sources
| Ureteric Tip |
| Ureteric Stalk |
| Ureter |
| Testis |
| Renal Vesicle |
| Proximal Tubule |
| Podocytes |
| Placenta |
| Ovary |
| Nephron Progenitor Cells |
| Muscle |
| Metanephros |
| Mesangial Cells |
| Lung |
| Liver |
| Kidney Tubules |
| Kidney Cortex |
| Kidney |
| Hypothalamus |
| HK-2 Cells |
| Hippocampus |
| HEK-293T Cells |
| HEK-293F Cells |
| HEK-293 Cells |
| Heart |
| Head Kidney |
| Epithelial Cells |
| Endothelial Cells |
| Cortical Collecting Duct |
| Colon |
| Cerebellum |
| Cap Mesenchyme |
| Breast |
| Brainstem |
| Brain |
| Bladder |
| Aorta |
| Adrenal |
| Adipose |
Biotypes
| 3prime_overlapping_ncrna |
| antisense |
| bidirectional_promoter_lncrna |
| IG_C_gene |
| IG_C_pseudogene |
| IG_D_gene |
| IG_D_pseudogene |
| IG_J_gene |
| IG_J_pseudogene |
| IG_LV_gene |
| IG_pseudogene |
| IG_V_gene |
| IG_V_pseudogene |
| lincRNA |
| macro_lncRNA |
| miRNA |
| misc_RNA |
| Mt_rRNA |
| Mt_tRNA |
| nonsense_mediated_decay |
| non_coding |
| non_stop_decay |
| polymorphic_pseudogene |
| processed_pseudogene |
| processed_transcript |
| protein_coding |
| pseudogene |
| retained_intron |
| ribozyme |
| rRNA |
| scaRNA |
| sense_intronic |
| sense_overlapping |
| snoRNA |
| snRNA |
| sRNA |
| TEC |
| transcribed_processed_pseudogene |
| transcribed_unitary_pseudogene |
| transcribed_unprocessed_pseudogene |
| translated_processed_pseudogene |
| translated_unprocessed_pseudogene |
| TR_C_gene |
| TR_D_gene |
| TR_J_gene |
| TR_J_pseudogene |
| TR_V_gene |
| TR_V_pseudogene |
| unitary_pseudogene |
| unprocessed_pseudogene |
| vaultRNA |
Metadata
| age | Adult |
| age | E10.5 |
| age | E11.5 |
| age | E12.5 |
| age | E13.5 |
| age | E14 |
| age | E14.5 |
| age | E15.5 |
| age | E16 |
| age | E16.5 |
| age | E18 |
| age | P1 |
| age | P4 |
| age | Weeks 14 |
| age | Weeks 15 |
| age | Weeks 16 |
| age | Weeks 17 |
| age | Weeks 24 |
| age | Weeks 3 |
| age | Weeks 42 |
| age | Weeks 5.5 |
| age | Weeks 6 |
| age | Weeks 8 |
| age | Weeks 9 |
| age | Years 19 |
| age | Years 22 |
| age | Years 29 |
| age | Years 37 |
| age | Years 45 |
| age | Years 47 |
| age | Years 51 |
| age | Years 60 |
| age | Years 65 |
| age | Years 67.5 |
| age | Years 68 |
| age | Years 73 |
| age | Years 77 |
| sex | Female |
| sex | Male |
| strain | 129X1/SvJ |
| strain | African |
| strain | BALB/c |
| strain | C57/BL6 |
| strain | C57BL/6 |
| strain | C57BL/6J |
| strain | C57BL6/6 |
| strain | Caucasian |
| strain | CD-1 |
| strain | DBA/2J |
| strain | DBAxC57BL/6J |
| strain | Singapore |
| strain | Six2-TGC |
Venn Diagrams
Clicking the [VENN] button will take the users to the view with Venn Diagram. This view allows the users to combine several searches, which are similar to the search in the LOCI view.Logic Programming
RenalDB is unique in that it uses logic programming when determining expression specificity. This allows samples at different levels of anatomical hierarchies to be compared. For example allowing whole organs, organ sub-tissues, and even component cell types to all be compared at once. The knowledge bases used in RenalDB can be found below, implemented in pyDatalog.Record Pages
Record pages can be accessed from the loci view. Record pages contain detailed information about a sequence contained within RenalDB. Record pages include annotation data, genomic coordinates, expression data, homologs, and GO terms depending on the sequence type.Visualization of Expression Data
There are two ways to view expression data: a graphical view and a table view. Users can toggle between the two views using the provided buttons.The graphical view features a tree showing the organization of the expressed samples and the heatmap showing the average magnitude of expression. In the heatmap darker values indicate stronger expression and lighter values weaker values. The tint of the heatmaps is determined using the equation y = 100.0 - (log(x)+12)*((3*x)/(x+1)+1), where y is the lightness value in the HSLA color.
Expression data is represented as between-sample normalized counts, divided by effective length, and finally multiplied by 10^3 to provide an FPKM like value. Note: FPKM could be obtained by dividing by the number of reads per sample then multiplying by 10^6. We excluded this step, in RenalDB as between-sample normalization was already preformed.
